Proceedings of the International AAAI Conference on Web and Social Media · 2021 · 94 citations · 0 references
Abuse DetectionEngineeringSocial Medium MonitoringInformation ForensicsPolitical BehaviorCommunicationJournalismText MiningNatural Language ProcessingComputational Social ScienceTwitter Political ActorsSocial MediaData ScienceSocial Medium NewsPolitical CommunicationContent AnalysisSocial Network AnalysisSocial Medium MiningWindow DressingKnowledge DiscoverySocial ComputingSocial BotnetUkrainian Social BotnetSocial Medium DataArtsEmpirical EvidencePolitical Science
A considerable amount of data in social networks like Twitter is not generated by humans but by automatic programs (bots). Some of these bots are mimicking humans (socialbots) and can hardly be identified. In this article, we analyze a social botnet involved in the Ukrainian/Russian conflict. Based on text mining and unsupervised learning, we can identify three different behaviors: mimicry, window dressing, and reverberation.